Optimising complementary soft tissue synchrotron X-ray microtomography for reversibly-stained central nervous system samples

Synchrotron radiation microtomography (SR mu CT) is a nominally non-destructive 3D imaging technique which can visualise the internal structures of whole soft tissues. As a multi-stage technique, the cumulative benefits of optimising sample preparation, scanning parameters and signal processing can improve SR mu CT imaging efficiency, image quality, accuracy and ultimately, data utility. By evaluating different sample preparations (embedding media, tissue stains), imaging (projection number, propagation distance) and reconstruction (artefact correction, phase retrieval) parameters, a novel methodology (combining reversible iodine stain, wax embedding and inline phase contrast) was optimised for fast (similar to 12 minutes), high-resolution (3.2-4.8 mu m diameter capillaries resolved) imaging of the full diameter of a 3.5 mm length of rat spinal cord. White-grey matter macro-features and microfeatures such as motoneurons and capillary-level vasculature could then be completely segmented from the imaged volume for analysis through the shallow machine learning SuRVoS Workbench. Imaged spinal cord tissue was preserved for subsequent histology, establishing a complementary SR mu CT methodology that can be applied to study spinal cord pathologies or other nervous system tissues such as ganglia, nerves and brain. Further, our 'single-scan iterative downsampling' approach and side-by-side comparisons of mounting options, sample stains and phase contrast parameters should inform efficient, effective future soft tissue SR mu CT experiment design.